The Nuts And Bolts of Advertising Online

Somebody really has to step up and lay down the rules for online advertising — What exactly constitutes an impression? What is a click? And so on.

That’s what I told myself about six months ago, after haggling all day with media vendors about their numbers not matching ours. Guess what I found out? Someone finally did. Well, sort of .

In The FAST Lane

In response to the call for reliable measurement, a group called the FAST Steering Committee emerged in the summer of 1998 with the mission to standardize online advertising.

Initiated by Procter & Gamble’s Future of Advertising Stakeholders Summit last August, the FAST coalition brought together several major industry organizations, including the IAB, the ANA, and the ARF. In September, advertisers, agencies, and online publishers met again at @d:tech New York, to assemble the FAST action items for 1999. Those became recommended guidelines for all of the following:

  • Consumer privacy
  • Ad models and creative formats
  • Audience and ad impression measurement
  • A standard reporting template, including insertion orders

In March, FAST announced completion of the guidelines for audience and impression measurement, including a complete technical explanation. They were the first to admit that “this guideline is the next step, and not the final step in the continuous process to establish a viable and useful advertising currency for the WWW.” According to the FAST “Basic Advertising Measures,” the metrics of online advertising are as follows:

Impression: “The count of a delivered basic advertising unit from an ad distribution point.”

Layman’s Translation: When an ad is “called up” from the advertiser’s server.

Problem: This definition only specifies that the ad has been “called up,” not that it has been fully loaded. Something as seemingly minute as that distinction can actually make for great differences between the numbers that two different servers count. Publishers prefer to define an impression as an ad being “called up.” Advertisers, on the other hand, would prefer that an impression be counted when the ad has fully loaded.

Click: “When a visitor interacts with an advertisement.”

Layman’s Translation: When the ad is clicked on and the visitor begins transferring to the destination site.

Problem: This definition does not count a click as someone actually arriving at the destination site. It simply defines a click as “the opportunity” to be transferred. Again, depending on whether you’re an advertiser or a publisher, this distinction can yield very different numbers.

Problems aside, I still think FAST is on the right track by attempting to lay the groundwork for online ad measurement and trying to inspire confidence in the traditional advertising circles. But FAST will still have to tackle the following major issues as they move forward:

  1. Few people are aware of FAST, including advertising decision-makers.
  2. “Voluntary recommendations” are not laws, and therefore do not require action.
  3. FAST has the difficult job of remaining unbiased as well as accurate in its recommendations. Not only do organizers have to walk the thin line between advertiser and publisher, they must also be very careful not to jump the gun. With the online media industry still so new, it may be premature to make laws before we know where the technology is ultimately going.

Let’s not lose sight of the bigger picture. The technology is in place right now to accurately track online advertising. The discrepancies really depend on how you define the metrics, which, of course, depends on who you are.

Sure, you can argue that caching, spiders, and other creepy things in the night will cause a glitch now and then. But let’s take a moment to stop and look at how other forms of advertising are measured. Random surveys? Sampling of the population? Tell me there isn’t room for error there. Yet no one seems to have a problem buying traditional media. (A silent teardrop falls. )

Here are my recommendations for advertisers until FAST is unanimously supported or until legislation is passed (neither of which will happen for an eternity in terms of web time):

  1. Find a third-party server which defines impressions and clicks the way you want them to be defined.
  2. Compare apples to apples on your own counts, so that you can get an accurate read of what’s going on. In other words, don’t rely on the individual sites’ numbers. Rely on your own numbers.
  3. Protect yourself far before invoice time. Put a sentence like this one at the bottom of every insertion order: “The client will be invoiced for the delivery of impressions that can be verified through only. If there is more than 5 percent discrepancy between the impression numbers by and the site, a make-good for the remaining impressions will be issued.”

Let’s remember that impressions and clicks are not the end-all-be-all of measuring an online advertising campaign. In fact, they’re just the beginning. It’s the post-click analysis that really counts.

In other words, the only thing that really matters to the client is the final ROI. FAST could jump off a cliff tomorrow, and it wouldn’t change a thing about the way I decide whether or not to keep a site or give it the axe.

Beyond The Click

Here’s an example of a weekly “by site” report that we use at M2K for our media clients

Click-through Rate: The number of clicks as compared to the total number of impressions served. The current industry average is speculated to be around .15 percent (according to NetGravity, March). This metric is important for: (1) making sure your creative is doing its job, and (2) being able to tell whether a site is giving you enough clicks for the price. In the above example, Site A seems to be getting a better overall click-through than Site B. It can also be inferred that on Site B, the Entertainment section is much more likely to get clicks than the Sports section.

Leads and Sales: These can be tracked a number of ways. But the most common is to put a hidden tag (1×1 pixel) on the “thanks for purchasing” page, and then track the hits to that tag. If this is out of your league, services like DoubleClick’s Spotlight or AdKnowledge’s Advisor can help you do this.

Cost per Lead and Cost per Sale: These are by far the most important numbers to consider when optimizing your campaign. Note above that Site A may have had higher click-through rates overall, but its cost per sale was much higher than Site B. Also, when analyzing Site B alone, it appears that the Sports section should be cancelled, and the money shifted instead into the Entertainment section.

That’s just a peek at the kind of in-depth analysis currently being used by the top online agencies. You still have to use your head, but it sure takes out a lot of the guesswork. And clients are beginning to catch on to this, realizing they can demand a higher level of accountability across all their advertising efforts as the technology evolves.

According to Rich LeFurgy, head of the FAST Steering Committee, “This is not sexy stuff, but it is the building blocks of how we do business It’s time for us to link arms and say, ‘The medium is good, and it’s going to get better.’ ”

Currently over 600 industry leaders have volunteered to take part in FAST. For more info, check out the FAST web site.

Related Links:

Fast Summit Media Links

Fast Summit Reviews:

ClickZ, Ann Handley: Day 1

ClickZ, Andy Bourland: Day 2

Internet World, Nelson Wang: “Fast Summit Follow-Up”

Business 2.0 April ’99 (print version),James Ryan “Not so FAST”

Dana Blankenhorn: “The 5-Second Rule for Internet Advertising”

Related reading

screen-shot-2016-09-13-at-10-20-04
mike-andrews
rsz_adblock
/IMG/224/276224/adblock-plus-logo-320x198
<